Hello Product Hunt 👋 I’m Abdesslem, the maker of Truvec. I built it because I kept seeing teams struggle when migrating from one embedding model to another. Truvec started as a simple way to compare embedding models on your own documents. It now goes further: 🔍 Compare retrieval strategies side by side on your real data 🐞 Debug failing questions to see exactly why the right chunk wasn't retrieved ⚙️ Experiment in a few clicks: add a reranker, change your chunking, swap the embedding model, or turn on BM25 hybrid search, then see how each change affects the results for every question The goal is simple: you shouldn't have to guess whether a RAG change helps or hurts. You should be able to measure it. I’d love your feedback, especially if you’ve been through an embedding migration or are tuning a RAG pipeline right now. What would make Truvec more useful for you? Thanks for checking it out! 🙏

Truvec

Find out why your RAG retrieves the wrong answers

Find out why your RAG retrieves the wrong answers